(November 25, 2011) Radiation exposure may have been underreported in the 1950s and ’60s, making it appear more deadly.
Low levels of radiation measurably increase deaths from cancer. Or so concluded a 15-country study into nuclear safety before scientists discovered that bad data from Canada had almost certainly corrupted its findings.
The embarrassed Canadians are now trying to make amends through reanalyses of the erroneous bits of data that they had supplied to the study’s researchers, and through reassurances that low levels of radiation are not all that dangerous. The Canadian redos will be helpful but inadequate, because the 15-country study itself, though the world’s largest ever into nuclear safety, was conducted under the premise that low levels of radiation carry some small risk. For unbiased, comprehensive results, the Canadians should open their eyes — and the eyes of their colleagues around the world — to what the data could really be saying: that low levels of radiation are not only safe, they may actually decrease deaths.
The blue-chip study, conducted by the World Health Organization’s International Agency for Research on Cancer, tracked more than 400,000 nuclear industry workers who during their careers wore a dosimeter, or badge, to record their exposure to radiation. For each of the badged workers who had died, the study determined the cause of death to determine if workers with higher radiation doses had a higher risk of cancer.
The initial finding: Low levels of radiation did indeed cause cancer — 2% of the 6,519 cancer deaths, and 1% of the 196 leukemia deaths that the population of 400,000 had experienced were statistically attributed to their radiation exposures.
But the scientists also noticed something odd — all of those excess statistical cancers were attributable to Canada. The other 14 countries found no statistically significant risk of cancers. Odder still, all of the excess cancers from Canada came from one of four data sets, the oldest data set, involving those who had worked for Atomic Energy of Canada Limited between 1956 and 1964. This one group, according to the data, were nine times likelier to die from cancer than workers who weren’t exposed to radiation. Without that old data set of early radiation workers, the data from the 14 countries and Canada’s other three data sets again showed no statistical risk of cancers.
Canada’s Nuclear Safety Commission then zeroed in on the early workers from the 1950s and 1960s, a period when workers were exposed to more radiation because safety standards were less stringent and because nuclear workers were more cavalier in their handling of nuclear materials. The CNSC all but concluded that the data were misreported in that nuclear macho era, possibly because workers didn’t always bother reporting the doses they received. If much radiation went unreported, the radiation that was reported would seem to be much more deadly.
The CNSC is now trying to pin down exactly how the data were collected. It also notes the curious detail that, even though that early group of Atomic Energy of Canada Ltd. workers suffered more cancer deaths than workers who weren’t exposed to radiation, they outlived the general population, dying less often from cancers as well as from non-cancer diseases. “This fact reinforces CNSC concerns that there remains a data problem as opposed to a true increase in their risk of solid cancer mortality,” the CNSC states.
Quite right. But this fact should also recall for the CNSC the parallel experiences of early radiologists in the U.K.
In the early part of the 20th century, when the U.K. radiologists had lax safety procedures, their mortality rate from cancer was 75% higher than that of all male physicians. Yet they weren’t dying in droves. To the contrary, U.K. radiologists had comparable death rates to those of other physicians because the radiologists’ extra cancer deaths were cancelled out by fewer deaths from other causes, such as heart and cardiovascular disease.
When radiation standards were tightened, so that radiologists received small doses of radiation, but more than physicians as a whole, radiologists died less often from cancers and also less often from other diseases. These findings came from none other than Sir Richard Doll, the Oxford epidemiologist best known for linking smoking with lung cancer in the 1950s. His 2003 study of British radiologists showed that those who entered the profession between 1955 and 1970 had a 29% lower risk of cancer (albeit not statistically significant) and a 32% lower death rate from all causes (statistically significant) compared with other physicians.
These findings mesh with those from numerous other studies that point to a sweet spot at which radiation becomes a help rather than a harm to human health. The bad data from Canada that corrupted the World Health Organization study would have an inadvertent benefit if it served to highlight the potential of low levels of radiation to promote human health, and if it then led Canada’s health authorities to confirm it or to disprove it.
This article first appeared in the Financial Post.